The exemplary embodiment relates to natural language incident resolution for a document processing device. While the systems and methods described herein relate to incident resolution using natural language descriptions, it will be appreciated that the described techniques may find application in other natural language processing applications, for networked computing devices, complex machinery, fault reporting, and the like.
In support of various devices, incident reports or tickets are generated indicating the nature of the support issue, a party assigned to correct the issue, and the like. These reports or tickets may be input automatically from a device reporting a fault to a monitoring service, a telephone or electronic mail message submitted by a customer to a physical help desk, via a web-based interface, or the like. In each of the non-device reporting circumstances, the customer initiated reporting relies upon a human describing the problem associated with the device using human natural language. After the problem is described, the personnel associated with the help desk are tasked with interpreting the problem based upon this description and then acting upon it.
For example, when a customer or user submits an incident, (either via text form input or via a phone call) the user describes the problem using natural language. The support personnel triaging the incident must manually correlate the customer's language to the current machine state or machine state history. The time and effort this manual correlation takes is time and money that could be used elsewhere or captured and assessed to customers. For example, the device may report the current machine state to the support personnel. When the issue is noticed by the customer, the customer may submit the incident via text or telephone call, to the support personnel. The support personnel must then manually link the submittal from the device with the submittal by the customer, so as to avoid having duplicate submittals requiring action on the part of the support personnel.
Accordingly, there is an unmet need for systems and methods that facilitate reduction of duplicate incident reports and automated correlation of device reported and customer reported incidents.